At present, a camera with an infrared function (e.g., a speed dome camera) uses an infrared light fill strategy, i.e., multiple groups of infrared lamps with different illumination angles are time-sharingly opened in different magnifications, so as to achieve an intention that video surveillance can be implemented in a low light environment, and a surveillance image is clear.
In addition, an average brightness value of the surveillance image can be obtained, the obtained average brightness value can be compared with a preset brightness reference value, and the brightness of the infrared lamp can be adjusted according to a comparison result, so as to achieve an intention of adjusting the brightness of the surveillance image.
However, according to the method above, the average brightness value of the whole surveillance image is used as an adjustment parameter. Thus, only rough adjustment for the whole surveillance image can be implemented, and it is not ensured that there is no brightness sudden change in each region in the surveillance image. That is, a certain sub-region may be overexposed or a too dark, so as to reduce image quality of the surveillance image.